Children and adults both see ‘pirates’ in ‘parties’: letter‐position effects for developing readers and skilled adult readers

Developing readers often make anagrammatical errors (e.g. misreading pirates as parties), suggesting they use letter position flexibly during word recognition. However, while it is widely assumed that the occurrence of these errors decreases with increases in reading skill, empirical evidence to support this distinction is lacking. Accordingly, we compared the performance of developing child readers (aged 8–10 years) against the end‐state performance of skilled adult readers in a timed naming task, employing anagrams used previously in this area of research. Moreover, to explore the use of letter position by developing readers and skilled adult readers more fully, we used anagrams which, to form another word, required letter transpositions over only interior letter positions, or both interior and exterior letter positions. The patterns of effects across these two anagram types for the two groups of readers were very similar. In particular, both groups showed similarly slowed response times (and developing readers increased errors) for anagrams requiring only interior letter transpositions but not for anagrams that required exterior letter transpositions. This similarity in the naming performance of developing readers and skilled adult readers suggests that the end‐state skilled use of letter position is established earlier during reading development than is widely assumed.     

Perceptual Categorization: Connectionist Modelling and Decision Rules

Although it is currently popular to model human associative learning using connectionist networks, the mechanism by which their output activations are converted to probabilities of response has received relatively little attention. Several possible models of this decision process are considered here, including a simple ratio rule, a simple difference rule, their exponential versions, and a winner-take-all network. Two categorization experiments that attempt to dissociate these models are reported. Analogues of the experiments were presented to a single-layer, feed-forward, delta-rule network. Only the exponential ratio rule and the winner-take-all architecture, acting on the networks' output activations that corresponded to responses available on test, were capable of fully predicting the mean response results. In addition, unlike the exponential ratio rule, the winner-take-all model has the potential to predict latencies. Further studies will be required to determine whether latencies produced under more stringent conditions conform to the model's predictions.